Hugging Face teams up with Google Cloud

If you've ever tried to pull an open-source model off Hugging Face and stand it up in a production cloud environment, you know the plumbing is rarely fun.

Google Cloud and Hugging Face just announced a major strategic partnership to make that workflow practically seamless. The deal integrates the premier hub of open-source AI directly into Google's cloud and hardware ecosystem.

With over 1 million models, datasets, and apps hosted on Hugging Face—and hundreds of thousands of developers already running them on Google Cloud every month—the integration targets a massive active audience.

Hugging Face meets Vertex AI and TPUs

Under the partnership, Google Cloud customers can train and deploy Hugging Face models directly inside Vertex AI and Google Kubernetes Engine (GKE).

Teams can deploy models on Google's top-tier compute options, including A3 VMs packed with NVIDIA H100 GPUs, C3 instances powered by Intel Sapphire Rapids CPUs, and Google's custom TPUs.

Developers hanging out on Hugging Face Hub get perks too. They can push models to production using Inference Endpoints backed by Google Cloud, and tap TPU compute directly within Hugging Face Spaces.

For enterprise teams, Hugging Face Enterprise Hub billing can now be routed directly through their Google Cloud accounts, cutting down corporate procurement friction.

Why it matters

For builders, this is a major quality-of-life upgrade. If your team wants to customize an open-source model rather than rely on closed APIs, you now have a frictionless path from finding a checkpoint to serving it at scale on enterprise infrastructure.

For Google, this is a very sharp tactical move. Microsoft has a massive enterprise lead with OpenAI, so Google is betting heavy on becoming the undisputed infrastructure layer for the open-source movement. Hugging Face users don't need to adopt Google models; Google just wants to sell them the compute to run whatever they want.

We think Hugging Face is playing this smartly by playing the field. Following its existing alliance with AWS, adding Google Cloud ensures Hugging Face remains the Switzerland of AI, keeping developers plugged into open models regardless of which cloud they choose.